Antonio Neri, president and CEO of Hewlett Packard Enterprise, wrapped up his keynote on the first day of HPE Discover 2023 by saying that this year's event was "the best HPE Discover in company history." Whether it was the best, time will tell. But it has a strong case for being the boldest.
Among the numerous announcements, one stood above the rest: HPE is entering the AI public cloud provider market -- not private cloud, but public cloud.
The initial offering is HPE GreenLake for large language models (LLMs), and it will:
- Enable enterprises to privately train, tune and deploy large-scale AI on HPE's Cray supercomputing technology offered as a public cloud service.
- Be delivered in partnership with German AI startup Aleph Alpha to provide users with a ready-to-use LLM to support use cases requiring text and image processing and analysis.
- Be deployed within sustainable colocation facilities, with QScale in North America serving as the first region.
- Be the first in a portfolio of AI public cloud services.
While HPE Discover featured other announcements, HPE's entry into the public cloud space with AI services is quite notable. Fortune favors the bold, and I applaud this bold and potentially risky business move.
HPE trusts the supercomputing technology and expertise that it acquired from Cray as well as its expertise in helping organizations on their AI journeys. That being said, HPE will need to prove that it can effectively manage and scale public cloud resources while delivering services that meet and exceed expectations. This is a competitive space with strong incumbents, and it will likely become even more competitive very quickly.
While HPE will face competition, such competition in AI public cloud services is great for users. Competition improves experiences, drives down prices and increases accessibility. For businesses and users, this is a huge win. Every organization is, or should be, developing its own AI strategy right now. What HPE offers is an opportunity to develop that AI strategy without requiring a massive capital outlay to buy and deploy infrastructure.
Will HPE be able to fully utilize its Cray supercomputing technology and expertise to lead the industry in AI public cloud services? It is still far too early to predict, but it is safe to say that every organization should -- in addition to developing an AI strategy -- be investigating what HPE is doing.
An effective AI strategy will be essential to staying competitive in the years ahead. In addition, research from TechTarget's Enterprise Strategy Group shows there isn't a one-size-fits-all approach to AI infrastructure. Artificial intelligence workloads were toward the top of the list among organizations that decide to keep a workload on premises, as 29% of organizations said AI/machine learning workloads were not a candidate for cloud migration. Others opt to use a secondary cloud provider rather than their primary cloud provider, as 31% of organizations with a strategy of prioritizing a primary cloud provider identified AI/ML workloads as driving the use of an alternative cloud provider. Those results do not mean that any particular location is unsuitable. Instead, it means that those data sets and workloads are so valuable that businesses apply greater scrutiny to their deployment. And the right deployment location can vary. Success in AI requires options, and HPE is introducing an intriguing new option for AI workloads.
Other HPE Discover announcements
Beyond the AI-related announcements, other major announcements at HPE Discover include:
- Multiple improvements in the integration between AWS and HPE GreenLake, including support for Amazon EKS Anywhere being made generally available on HPE GreenLake for Private Cloud Enterprise. Nearly every organization is multi-cloud now, and integration is essential.
- Multiple enhancements to HPE GreenLake along with a new offering, HPE GreenLake for Private Cloud Business Edition. This offers an easy method of deploying VMs across hybrid cloud environments.
- An expanded partnership with Equinix, through which HPE will pre-provision HPE GreenLake for Private Cloud Enterprise and HPE GreenLake for Private Cloud Business Edition at seven Equinix data centers around the world. This move should accelerate HPE's ability to provision HPE GreenLake services within just a few days, significantly improving time-to-value.
- The close of the OpsRamp acquisition, which offers some impressive capabilities, including discovery, monitoring, AIOps-based remediation and automation, for multi-vendor, hybrid and multi-cloud environments. Its monitoring capabilities can go up the stack to VMs and to databases and apps as well. HPE also leverages OpsRamp's heterogeneous capabilities to significantly increase the scope of its sustainability dashboard, extending its capabilities to include third-party infrastructure for a more complete measure of an organization's environmental impact. HPE said it plans to continue to offer the SaaS offering option, in addition to proving the OpsRamps via HPE GreenLake as a new service.
HPE continues to build out HPE GreenLake with more services and more value.
For HPE's current and potential customers, AI strategy is so essential to future business success that no stone should remain unturned when evaluating options. AI is not a workload that should be cloud-by-default, or even on-premises by default. Businesses must put in the work to figure out what is the right option for them, and many are already doing that.
This gives HPE an opportunity. If it can prove it can deliver the full benefits of its supercomputing technology and expertise in a manner that is easier, faster and more cost effective, then that feat would fundamentally transform the way the world views HPE in five years' time.
Enterprise Strategy Group is a division of TechTarget. Its analysts have business relationships with technology vendors.